We had a great talk and discussion within our RPA community during the last two series of Microbiome Friday Seminars!
Alejandro Rincon gave an introductory talk on machine learning approaches for clinical data during the seminar held on March 7th. He shared the application of recursive ensemble feature selection on microbiome data to predict various conditions, such as autism spectrum disorder, inflammatory bowel disease (IBD), and type 2 diabetes. Exciting work!
The afternoon continued with a talk by Gonçalo Piedade on antimicrobial resistance in infants from a longitudinal cohort study in Zambia. He also shared his observations on the specific species of bifidobacteria found in the gut of babies in Zambia during the weaning period. The discussion with participants extended to the role of breastfeeding in the antimicrobial resistance profile in early life.
On April 4th, we invited Indra Bergval from RIVM to give a talk on microbiome analysis of food products. She shared her exciting research on microbial ecosystems within the food production chain. Specifically, she demonstrated the use of metagenomics and amplicon sequencing data to analyze the microbial community of raw chicken products from supermarkets.
In the second talk, Roel van der Ploeg presented his innovative approach to using multi-way methods for microbiome data analysis. He highlighted three compelling studies where these methods (PARAFAC, NPLS, or ACMTF) were successfully applied, particularly in longitudinal microbiome data. These studies included research on gingivitis, gender-affirming hormone therapy, and an apical periodontitis cohort. Check out these toolbox packages: on parafac4microbiome, NPLStoolbox, and CMFtoolbox, available on Roel’s github!
